Don't forget serendipity

If the new tools hold out the potential for improving the choice of targets and compounds, Icos Corp. Chairman George Rathmann cautioned about becoming too clever.

"The cynic in me says as I get smarter, and as people doing research get smarter, there's more awareness of risk and you get so sophisticated you start asking every way things can go wrong - is there a patent issue, is there a competitor? For example, we have a molecule that destroys PAF - and British Biotech plc had a receptor antagonist to block the action of PAF. Should we continue with our program? If I'm looking not to spend money, what a great opportunity to can a program: British Biotech was in the clinic, they were making progress. But eventually that approach failed, so we've got a breath of fresh air.

"I'm of the mind that a little more naivete gets you a better result. The most successful people have a way of not being objective about their field. Too much sophistication is a quick route to abandoning programs.

"We ask all the time - could we eliminate one-third, one-half, 10 percent, 70 percent of what we're working on by eliminating the three weakest programs and selecting the best three and rescrutinizing in six months. But the risk is that for reasons that were not predictable, even with the best three programs the probability of success is not greater than 33 percent. So the risk is not zero. Focusing is very easy. Bu you can get shot down for reasons you can't predict. Animal data can look terrific and in humans you can get nothing."